A number of children have difficulty with language learning despite what appears to be normal development in other areas (e.g., neuromotor, cognitive, social). These children referred to as specifically language impaired (SLI), are the focus of this research. Some SLI children appear to have age-appropriate lexical comprehension, but are considerably behind their peers in their ability to express their ideas through language (these children will be referred to as SLIexp). The specific aim of this project is to examine the possibility that the problem of SLIexp children is with encoding speech responses (i.e., forming phonetic representations and motor programs) as opposed to other operations (e.g., perceptual encoding or retrieving semantic information). The long term objective of this research is to increase our understanding of the language-learning problems of SLI children so that we may be better able to facilitate their language development. Performance of SLIexp and non-language-impaired (NLI) children, aged 4 to 8, will be compared on a number of tasks including confrontation naming with a delay interval of varying durations before response, confrontation naming with a semantic prime presented for varying durations, auditory lexical decision, and the repetition of real words and nonwords. These tasks are designed to place different demands on perceptual encoding, accessing semantic information and encoding information for speech production. Responses on these tasks will be used to construct discriminant functions to differentiate among four different groups of children: SLIexp children, SLI children with deficits in both language production and lexical comprehension, NLI children matched to the SLlexp children on the basis of chronological age, and NLI children matched to the SLlexp children on the basis of expressive language skill. We predict that measures that reflect heavy demands on speech encoding will emerge as significant discriminators or SLIexp children from their NLI age peers, whereas measures that reflect demands on perceptual encoding and lexical access will discriminate SLIexp children from SLI children with deficits in both language production and lexical comprehension. This research differs from other studies in a number of ways. It focuses on one subgroup of SLI children and is designed to separate the relative influence of encoding for speech production from the influences of perceptual encoding and lexical access.